[ntp:questions] Re: Allan variance plots

Piotr Trojanek ptrojane at mion.elka.pw.edu.pl
Wed Sep 17 15:34:07 UTC 2003


In article <slrnbmgm2p.7ue.ptrojane at mion.elka.pw.edu.pl>, Piotr Trojanek wrote:
>I feed this data directly to dataplot package (NIST software) and use
>'AV PLOT' command in log-log scale -- result is straight line with
>slope +1, so now I think it is wrong method. I can't figure out the
>right way.

Another problem -- slopes:  from "Adaptive Hybrid Clock Discipline
Algorithm for the Network Time Protocol" (allan.pdf file), page 8:

"The phase variations, which are dominated by white phase noise,
produce straight lines with slope -1 on the plots. The frequency
variations, which are dominated by random-walk frequency noise, produce
straight lines with slope +0.5 on the plots."

According to DATAPLOT reference cards:

ALLAN VARIANCE:
  white noise: -0.5
	random walk: +0.5
ALLAN STANDARD DEVIATION:
  white noise: -1
	random walk: +1

so what is actually on the plots on page 8? AV or ASD?

-- 
Piotr Trojanek



More information about the questions mailing list